This document is the strategic foundation for the Provenance Lab's identity. It is not a style guide or a press kit — it is the thinking that precedes both. It answers: what do we exist to do, who do we serve, what is our singular claim, and how do we speak and behave consistently across all contexts.
Purpose, vision, mission, values. Who we serve, what category we occupy, how we differ.
Brand promise, tagline candidates, four message pillars, proof points.
Personality, tone, vocabulary, visual principles, outputs, teaching, artistic dimension.
Who speaks, what stays consistent, content pipeline, project sub-brand architecture.
Internal · Strategic · Single-page scroll deck · April 2026
Every auction catalogue entry, every change of hands noted in a museum file, every gap in the documentation: this is not administrative residue. It is the primary evidence of how violence, displacement, capital, and power moved through the art world across the 20th century and before.
The Provenance Lab reads this evidence at scale — using computational methods to ask questions across thousands of objects simultaneously — and with care, to ensure the individual biography, the specific person, the particular moment of loss is never flattened into a data point.
Most provenance research produces findings that live in paper reports, institutional archives, or academic journals. We produce linked open data — machine-readable, openly licensed, interoperable — so that findings persist, accumulate, and become usable by other researchers, institutions, and legal bodies long after any single project ends.
We care about outcomes — restitution, reparation, honest institutional accounting — but our research is not in service of any predetermined outcome. Scientific independence is the condition of being trusted by all parties. We are useful to repair precisely because we are not the property of any side.
A world in which the ownership history of every significant cultural object is legible, machine-readable, and useful — to researchers tracing displacement, to institutions seeking accountability, to communities pursuing justice. Not a perfect world, but one in which the gaps in the record are documented honestly rather than papered over.
We reconstruct ownership histories of thousands of artworks using a combination of archival research, artificial intelligence, and linked open data. We publish our data openly under Creative Commons. We build tools — PROV-A, NLP annotation models, linked data schemas — that other researchers and institutions can use. We train researchers at the intersection of art history, computational methods, and the ethics of provenance.
artworks with provenance data across 36 US museums — Modern Migrants alone
Germany's first Lichtenberg Professorship dedicated to provenance studies, Leuphana
Provenance research must be emancipated from political pressure and restitution expectations. Not because outcomes don't matter — they do — but because science captures by an interest is science that cannot be trusted by anyone else. Independence is the precondition of credibility.
Computational methods allow us to ask questions across thousands of objects. Humanistic methods ensure we answer them with precision and moral seriousness. The tools serve the story. We never let scale be an excuse for flattening what is individual, specific, irreplaceable.
The histories we tell require art history, economic and legal history, sociology, and data science simultaneously. None is sufficient alone. The gaps between disciplines are exactly where the most important questions live — and where the Provenance Lab is designed to operate.
Vagueness, incompleteness, subjectivity, and uncertainty are not defects in provenance data — they are its honest condition. The VISU framework documents them explicitly. We do not smooth over gaps; we name them. In a field where false certainty can derail a restitution claim or bury an injustice, epistemic honesty is an ethical commitment, not just a methodological one.
Objects are not merely art. They are evidence of inheritance, displacement, wealth, persecution, and power. Behind every change of hands is a human story — a forced sale, a flight from persecution, a colonial extraction, a dealer navigating multiple worlds. We follow the objects to find the people. That is the animating commitment of everything we do.
Heirs of victims of Nazi persecution seeking the return of looted or forcibly sold works. Postcolonial communities, source nations, and Indigenous groups pursuing the repatriation of objects removed in the context of colonial extraction. These are the people whose lives the provenance record most directly concerns — and whose ability to make claims depends on the existence of honest, accessible, and legible documentation. The lab's open data and transparent methodology exist, in part, so that claimants are not the last to know.
Art historians, digital humanities researchers, economic and social historians, legal scholars working on cultural heritage. The lab produces methods, data, and tools they can build on. We train the next generation through doctoral supervision, the Coding Provenance Workshop, and published open-access datasets.
The MoMA partnership, Getty collaboration, and 36-museum Modern Migrants dataset demonstrate the model: we embed within institutions to access the records needed to produce honest, large-scale provenance analysis. Institutions benefit from the research; the research benefits from access to the collections.
The Volkswagen Foundation, Lower Saxony, the Getty — funders who understand that provenance knowledge infrastructure is generational work, not a project. We make the case for sustained investment in methods and data that compound over time.
The Return of Hatshepsut demonstrates the model: artists treating provenance as a form of inquiry, not just a documentary record. The lab is a research partner, not a service provider, in these collaborations.
Journalists covering looted art, restitution policy, colonial collections. Policymakers and legal professionals who need methodologically rigorous findings. The lab's scientific independence makes its data usable in contested contexts.
whatisapainting.com as a model: research findings made experiential without losing intellectual honesty. The public has a stake in knowing how collections were formed. We make that knowledge accessible.
Research agency using disciplinary methods as instruments of accountability. Independent, embedded in a university, builds tools. Key difference: FA is explicitly activist-political. The lab is scientifically independent — caring about outcomes without being captured by any of them.
Knowledge infrastructure at scale. Deeply influential, long-term. Key difference: the GPI is curatorial infrastructure within a major institution. The lab is critical, independent, and university-based — a research partner to the Getty, not an arm of it.
No other center combines all three. The lab's distinctiveness is exactly this triangulation — methodologically advanced, institutionally independent, morally serious. This is the category we own.
Not just linked data as a method — but provenance research whose outputs are published as open, machine-readable, interoperable data that other institutions can query, build on, and correct. This is what moves provenance from documentation to infrastructure.
Most institutions and research programs work in one domain or the other. The lab operates across both — recognising that the methodological and ethical challenges are structurally similar, and that a unified approach produces better science.
Germany is the epicenter of provenance reckoning. Being based at Leuphana means working in the most important national context for this discipline while operating globally — MoMA, Getty, 36 US museums, researchers from 40 countries.
Many labs use computational methods. Few pair them with the deep archival and art-historical training required to evaluate what the algorithms find. The lab does both — and the combination is not compromised; it is the point. Machine-scale pattern detection guided by humanistic judgment.
The lab's findings are usable in legal proceedings, policy debates, and restitution discussions precisely because they are produced by researchers with no financial or political stake in the outcome. In a field where every major actor has an interest, independence is a rare and valuable asset.
Lynn Rother's appointment as MoMA's first Curator for Provenance (2024–2026) is not incidental. It demonstrates what the lab's research program produces: a new institutional function that did not exist before.
The promise has two halves. The first is the technical, long-term, infrastructure half. The second is the humanistic, narrative, people-centred half. Neither is sufficient without the other. The lab is the place where both are true simultaneously.
Note on selection: The lab should choose one primary tagline for consistent use and hold the others as contextual variants. The preferred candidates above work across internal, funder, and public contexts. Final selection is a team decision.
We analyse thousands of artworks at once using artificial intelligence — pattern detection across provenance records that no human researcher could perform manually. But this scale is not an end in itself. It is a way of finding stories that would otherwise remain invisible: the single family whose collection moved through seven hands across three countries; the dealer whose name appears in a hundred records but has never been identified; the gap that repeats across a dozen museums and means something.
The tools serve the story. We never let computational scale become an alibi for treating people as data points.
Built to ensure that AI-processed provenance data does not produce false certainty. Every gap, every ambiguous attribution, every contested date is documented as such. Scale without epistemic honesty produces misleading results.
The most common criticism of provenance research is that it serves the interests of whoever funds it — institutions protecting their collections, or advocates pressing for maximal restitution. The lab's founding intellectual claim, developed in Rother's 2024 essay "Auf dem Weg zu einer emanzipierten Provenienzforschung," is that both capture modes corrupt the science and ultimately undermine the justice they claim to serve.
Independence is not indifference to outcomes. It is the structural condition that makes the lab's findings credible in adversarial contexts — legal proceedings, policy debates, restitution negotiations — where every other actor has an interest.
In field-theoretic terms (Bourdieu), provenance research has spent 25 years being instrumentalised by adjacent fields — law, politics, art history, journalism. The lab's project is to establish provenance research as a field with its own logic, standards, and capital. Not dependent on the outcomes it enables, but consequential because of the quality of the knowledge it produces.
Most academic research produces papers and conclusions that are archived, cited, and eventually superseded. The lab produces papers and infrastructure simultaneously: linked open data published under Creative Commons, tools like PROV-A that institutions can adopt, annotation schemes that standardise how provenance uncertainty is recorded, NLP models that other researchers can extend.
The distinction matters for funders: an investment in the lab is an investment in compounding resources, not a one-time project.
The Coding Provenance Workshop (20 participants, 11 countries, 200+ applicants from 40 countries) is infrastructure of a different kind: it is producing the researchers who will carry this methodology forward globally. When the lab trains a provenance researcher at the British Museum or the Van Gogh Museum in computational methods, that training does not expire.
countries represented in applications to the first Coding Provenance Workshop. The field is global; the lab is building its infrastructure.
An Impressionist painting acquired by a Berlin collector in 1912, sold under duress in 1935, purchased by a New York dealer in 1938, donated to a US museum in 1952 — this is not an art history story. It is a story about the collapse of Jewish bourgeois life in Germany, the mechanics of flight, the structure of the transatlantic art market, and the conversion of displaced cultural capital into institutional prestige.
The painting knows. The provenance record, read carefully, tells the story. The lab's task is to build the methods and infrastructure to read those records at scale — across 6,295 paintings, across 36 museums — and then to tell the story honestly.
This is the animating commitment that makes the lab more than a data science project, and more than a restitution instrument. It is a form of social and economic history conducted through objects.
Modern Migrants' core research question is not just "where did these paintings come from?" but "what does the aggregate movement of thousands of paintings tell us about the political economy of the 20th century?" Inheritance patterns, wealth distribution across generations, gender dynamics in collecting — these are visible in provenance data read at scale.
Colonial provenance data tells a parallel story: how objects were extracted from communities in the context of imperial expansion, how they moved through ethnographic and natural history museums, how the record was constructed to obscure the conditions of acquisition. VISU applies here too — uncertainty about origin is often not incidental but structural.
artworks with provenance data across 36 US museums. Modern Migrants, 2019–2027.
Germany's first and only Lichtenberg Professorship for Provenance Studies. Volkswagen Foundation, 2019.
University research program anywhere to produce provenance as linked open data.
Countries represented in Coding Provenance Workshop applicant pool. 200+ applications for 20 places.
Precise in language, generous in scope. Interdisciplinary without being eclectic — the connections we make are earned, not decorative.
We do not exaggerate. The facts are already remarkable. A ledger entry from 1937 contains more human consequence than most fiction. An auction catalogue from 1938 is a document of dispossession. We let that weight speak; we do not need to amplify it.
Citation-forward, range across Bourdieu and CIDOC-CRM, economic history and linked data standards. But always with a traceable argument. The claim is visible.
The object first. The person. The specific date. Then the pattern. Then the implications. Avoid jargon where a precise plain alternative exists. Do not talk down; talk accurately.
We have built something unusual. We can say so plainly, without false modesty or inflation. Specific numbers, specific firsts, specific named partners. No vague impact claims.
These are not buzzwords. They are terms the lab has developed or deployed with specific intellectual content. Using them consistently is part of how the brand maintains coherence across languages and contexts.
| Term | Meaning in our usage | Context |
|---|---|---|
| Emancipated provenance research | Research independent of political and restitution instrumentalization — free from capture by any party to the dispute | Academic writing; field-positioning |
| Provenance linked open data (PLOD) | Machine-readable, openly licensed, interoperable provenance as public infrastructure — not just digital, but queryable by others | Technical + funder communication |
| VISU framework | Vagueness, Incompleteness, Subjectivity, Uncertainty — the lab's framework for documenting epistemic limits in provenance data honestly | Methods papers; institutional communication |
| The social life of objects | Objects as carriers of human biography, economic force, political history — provenance as social and economic history conducted through things | Public communication; curatorial contexts |
| Reparative practice | The use of provenance knowledge in active service of repair — honest accounting as a form of institutional justice | Partnership and funder communication |
| Emancipated from / emancipatory for | The double register: free from political capture; generative for justice. Both true simultaneously. | Academic and policy contexts |
| PROV-A | The lab's provenance as linked open data management platform — an operational tool, not just a research output | Technical and partner contexts |
The primary visual material of the lab is its evidence: network graphs of ownership chains, maps of object migration, archival documents, auction catalogue pages, linked data diagrams. These are not illustrations — they are the argument. Design serves them, not the other way around.
A provenance gap is not beautiful. A forced sale in 1938 is not a design element. The lab's visual language is precise and restrained — clean enough to let the evidence carry weight, never styled in a way that converts historical injustice into aesthetic experience.
The lab produces materials that range from 19th-century auction catalogue transcriptions to CIDOC-CRM JSON-LD. The typographic system should be capable of handling both with equal authority — monospace for technical/data content, humanist sans for narrative and argument.
Research findings as experience. The Mondrian-style generative layout of whatisapainting.com takes one dataset finding (the inconsistency of museum definitions of "painting") and makes it visceral. This approach — data-driven, unexpected, intellectually honest — should be the model for public-facing outputs. Not infographics. Not explainer videos. Data made experiential.
The lab operates in both languages with equal authority. Neither is a translation of the other. Visual identity should support this — templates, document designs, and web presentation should function in both languages without one appearing to be a reduced version of the other.
A full visual identity system — typefaces, colour palette, logo treatment, document templates, web design system — is a second-phase deliverable. These principles are the brief for that work.
Every publication that can be open access, is. Every dataset is published under Creative Commons. Every methodological paper documents the VISU categories. The lab's research outputs are designed to be extended by others, not protected from them.
This is not just values — it is strategy. Open infrastructure builds a network of researchers who depend on and contribute to the lab's data layer. That network compounds the lab's influence faster than any individual publication.
PROV-A is the operational proof of the lab's infrastructure claim. It is not a research prototype — it is a tool that partner institutions can use to manage their own provenance data in linked open data format, interoperable with the Getty Provenance Index and Wikidata.
The linked data schemas, annotation models, and NLP tools built by the lab are designed for reuse. When institutions adopt them, the lab's methodology spreads — and the data that comes back enriches the common pool.
The first international workshop dedicated to computational provenance training. 20 researchers selected from 200+ applications from 40 countries. Participants from the British Museum, Musée du Quai Branly, Van Gogh Museum, National Gallery of Art.
This is not a conference. It is an immersive, method-focused training that the participants carry into their institutions. A second edition at the Getty is planned for end of 2026. This is how the discipline builds.
countries represented among 20 participants. Selected from 40 countries of application.
This project — a co-production between Marc Da Costa and Initiative BLACK LAND e.V., with the Provenance Lab as research partner — demonstrates that the lab's questions extend beyond data and into the register of art. The project interrogates the legacy of German and American archaeological excavations in Egypt and Syria at the turn of the 20th century, incorporating undisclosed archival documents and presenting them publicly for the first time.
Berlin, December 2026 (Kunstquartier Bethanien, as part of "Worlds on Repair"). New York, July 2027 (Onassis Foundation / The New School). With artists Yara Mekawei and Rayyane Tabet; curators Elena Sinanina and Sandeep Sodhi.
Provenance as artistic research is not peripheral to the lab's mission. It is another mode of the same inquiry — one that reaches audiences and produces understanding in registers that data and academic papers cannot.
The MoMA partnership, the Getty collaboration, the 36-museum Modern Migrants network — the lab embeds within institutions rather than criticizing from outside. This is deliberate. Access to primary sources in institutional archives is the precondition for honest provenance research. Proximity to institutional decision-making allows the lab's research to be consequential, not just accurate.
Embeddedness is not compromise. It is method. The lab maintains scientific independence precisely because it has formal relationships with multiple institutions simultaneously — no single institutional interest can capture it.
An interactive web visualisation presenting randomised museum definitions of "painting" through a dynamic, Mondrian-style generative layout. One finding from Modern Migrants — the radical inconsistency of how museums define the basic category of their collections — made tangible without dumbing down.
This is the model for future public-facing outputs: research-driven, visually unexpected, intellectually honest.
Prof. Dr. Lynn Rother is the primary voice in academic, institutional, and media contexts. The team — Dr. Max Koss, Dr. Bárbara Romero Ferrón, Fabio Mariani, Dr. Agustina Andreoletti, Coco Amy Rufer — are authors, co-presenters, and public voices in their own right. Every shared output carries the names of all contributors. The lab does not have one voice; it has a voice.
provenance-lab-hub is the institutional source of truth — every person, project, publication, press item, partner relationship, and strategic document lives here first. provenance-lab-web is the curated public presentation layer. All updates go into the hub first, then are propagated to the web. Never the reverse. This is what keeps the brand coherent as the institution grows.
Each project inherits the lab's intellectual identity. Some have their own public-facing identities; most operate under the lab brand directly.
| Project | Relationship to lab | Own identity? | Status |
|---|---|---|---|
| Modern Migrants | Flagship project — defines the lab's scale and primary research claim | Yes — modernmigrants.art | Active 2019–2027 |
| PAESE 3.0 | Colonial provenance sub-project, part of ProSaDi network | No — lab brand only | Active 2024–2028 |
| whatisapainting.com | Experimental public output from Modern Migrants research | Yes — standalone microsite | Live |
| Coding Provenance Workshop | Signature training program; co-branded with Getty Provenance Index | Co-brand: Lab + Getty | Annual; 2nd edition planned Getty 2026 |
| Return of Hatshepsut | Artistic research co-production; lab as research partner | Co-brand: Da Costa / BLACK LAND / Lab | In development; Berlin Dec 2026 |
| (Un)Mapping Infrastructures | International scholarly network; lab participates | No — external lead (U Amsterdam) | Concluded 2025 |
Make the hidden histories of art objects visible — and build the infrastructure that turns those histories into tools for scholarship, accountability, and repair.
Provenance is not a footnote. It is the primary evidence of how power, capital, and violence moved through art history. Reading provenance at scale is a form of social history conducted through objects.
A world where every significant cultural object carries an honest record of how it moved through human hands — machine-readable, openly licensed, and useful to those seeking justice.
Researchers, museums, funders, artists, and institutions that need to know the truth about how collections were formed.
The only research center that combines AI-scale analysis of provenance data with deep archival scholarship and formal scientific independence — first in Germany, operating globally.
Unlike legal services and advocacy programs, we produce science. Unlike purely digital humanities labs, we are grounded in the human consequences of what the data reveals. Unlike policy-driven provenance programs, our findings are not calibrated to any outcome.
To build the data infrastructure that makes repair possible — and to tell the stories that make it necessary.
Rigorous, curious scholars who do not exaggerate — because the facts are already remarkable.
This brand platform is not a rulebook. It is a shared framework — a way of ensuring that every paper, every workshop, every partnership, every website page expresses the same underlying logic without needing approval from a central authority.
When you are unsure whether something is consistent with the lab's identity: consult the platform. When the platform does not have an answer: the gap is an opportunity to extend it.
This document should be reviewed when:
Otherwise: annual review, April. Maintained in brand-positioning/brand-platform.md in the hub.
Provenance Lab · Brand Platform · 22 Slides · April 2026 · Internal · Not for distribution